Decomp/recomp scene explodes as frontier models crack ROM reverse engineering — can open-weight models keep up?
EmPips · reddit · 2026-10-08
The poster says closed models (Opus 5.5, Sol 6.1, Fable, Astra) have proven capable of full decompilation/recompilation projects, and the scene exploded this past week — partly thanks to maturing tools and feedback loops.
He asks whether any open-weight model (including K3 and Mimo-2.6) can do it, and whether larger community-run models like GLM 5.3-Flash, Qwen 3.8-Next-Flash, or V4.1-deepseek Flash could handle simpler targets like GBA and PSP with smaller ROMs and mature pipelines.
Decomp/recomp demands precise code understanding and long-horizon consistency — an interesting litmus test for coding models.
More from coding & agent
- Dev builds "Finger Contact" tool fixing floating fingers in guitar mocap using Guitar Pro tabs — TinfoilTricorn · 2026-10-08
- Ramp's hidden markdown-file offer read only by AI actually worked, customers claimed it — gaganghotra_ · 2026-10-08
- Creator revives a dead Notion course by feeding it to Codex as an interactive app — RichardsonDx · 2026-10-08
- Gary Bernhardt: AI 'factories' repeat the microservices over-engineering mistake — sergeykarayev · 2026-10-08
- Hierarchical RL with mixed discount rates may unlock long-horizon agent tasks — jessi_cata · 2026-10-08
- Graph-MIND: local MCP memory server hits 88.8% LongMemEval with zero LLM calls at write time — BrilliantGrocery8233 · 2026-10-08